A real-time control method for temperature and humidity of a multi-zone variable air volume air conditioning system

Abstract

The present invention discloses a real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system, comprising the following steps: (1) setting the room temperature and humidity values; (2) obtaining the temperature and humidity of the room through temperature and humidity sensors; (3) decoupling the room temperature and humidity by using a deep fuzzy cognitive map; (4) respectively comparing and taking the difference between the decoupled values of the room temperature and humidity and the set values. If the difference is not equal to 0, then taking the difference as the control error e(k) and inputting it into the RBF-PID algorithm; (5) obtaining the opening degrees of the terminal air valve and the chilled water valve at the next moment by using the RBF-PID algorithm; (6) controlling the controller to make the terminal air valve and the chilled water valve reach appropriate opening degrees; (7) repeatedly executing steps (2) to (6) and sampling at the sampling time period. On the basis of ensuring the terminal control efficiency, the present invention realizes the precise control of the room temperature and humidity, and has important theoretical significance for the development of the multi-zone variable air volume air conditioning system.

Description

Technical Field

[0001] The present invention relates to the field of operation regulation and control of variable air volume air conditioning systems, and particularly to a real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system. Background Art

[0002] Buildings are one of the three major economic sectors of a country and are the largest energy consumers. According to statistics, the energy consumption of buildings accounts for 40% of the global annual energy consumption, and the carbon emissions account for 36% of the total. Among them, the energy consumption of air conditioning systems is a major part of the total building energy consumption, mainly because the operating energy consumption of air conditioning systems remains high. Reducing the operating energy consumption of air conditioning systems has become the primary focus of building energy conservation while ensuring the thermal comfort of occupants.

[0003] The end air valve dampers of traditional VAV air conditioning systems are often controlled by proportional-integral (PI) algorithms or proportional-integral-derivative (PID) algorithms. The traditional PI algorithm cannot handle fixed parameters, which will cause frequent fluctuations in the valve opening and oscillations in the control system. The PID algorithm is a typical closed-loop control algorithm, which has the advantages of simple structure, easy implementation, and convenient adjustment. PID can calculate the adjustment amount of the valve opening according to the deviation between the measured value and the set value of the indoor temperature. The temperature and humidity in an air-conditioned room are not only related to the building envelope structure, but also have a great relationship with solar radiation, indoor heat disturbance, air supply volume, etc., and the indoor temperature and humidity are non-linear and coupled. If only the PID control method is used to control the valve opening, it will cause frequent changes in the valve opening, and it is difficult to achieve the control requirements of high-precision and high-performance for the valve opening. In order to solve the above problems, researchers combine neural networks, fuzzy control, predictive control theory with the PID control algorithm. In view of this, the present invention proposes a real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system, so as to effectively and real-time control the opening of the end air valve and the chilled water valve of the variable air volume air conditioning system. Summary of the Invention

[0004] The present invention aims to provide a real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system, so as to solve the technical problems of poor control effect and weak anti-interference ability in the existing multi-zone variable air volume air conditioning control system, so that the multi-zone variable air volume air conditioning system can perform linkage control on the indoor temperature and humidity, thereby improving the indoor air quality.

[0005] To solve the above technical problems, the present invention adopts the following technical solutions:

[0006] A real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system, comprising the following steps:

[0007] Step 1: Set the set values of the room temperature and humidity;

[0008] Step 2: Obtain the temperature and humidity of the room through a temperature and humidity sensor;

[0009] Step 3: Decouple the room temperature and humidity using a deep fuzzy cognitive map;

[0010] Step 4: Compare and subtract the decoupled values of the room temperature and humidity with the set values respectively. If the difference is not equal to 0, then use the difference as the control error e(k) and input it into the RBF-PID algorithm;

[0011] Step 5: The RBF-PID algorithm calculates the terminal air supply volume and supply air temperature at the next moment based on the difference, and obtains the opening degrees of the terminal air valve and the chilled water valve at the next moment;

[0012] Step 6: Input the data of the opening degrees of the terminal air valve and the chilled water valve obtained by the RBF-PID algorithm into the air volume controller and the supply air temperature controller respectively. The controller controls the terminal air valve and the chilled water valve to reach appropriate opening degrees;

[0013] Step 7: Loop through Steps 2 to 6 and sample at the sampling time interval.

[0014] Furthermore, the deep fuzzy cognitive map in Step 3 can decouple the room temperature and humidity according to the potential relationship between the room temperature and humidity:

[0015] 1) The state in the deep fuzzy cognitive map is affected by the states of other concepts at time t:

[0016]

[0017] where the function f j (·) is used to simulate the relationship from a to a j and is called the f-function, represents the system state of all concepts at time t, and the function u j (·) is used to simulate the influence of external factors on a j and is called the u-function;

[0018] 2) In the input layer of f j (a (t) ), assume in the output layer, an output can be obtained:

[0019]

[0020] where, is the output of the nth neuron in the Kth layer at time t, and υ(n1,K + 1) is the connection weight from the neurons in the Kth layer to the neurons in the K + 1th layer;

[0021] 3) The influence of external factors on a j can be indirectly measured by a u - function:

[0022]

[0023] 4) Let Then the deep fuzzy cognitive map defined in

[0024]

[0025] where, is the decoupling value of, θ f 、θ u are the parameters of f j 、u j respectively.

[0026] Furthermore, the θ f 、θ u are obtained by training the deep fuzzy cognitive map through the alternating function gradient descent algorithm:

[0027] 1) Input the training data set:

[0028] 2) Randomly initialize θ f 、θ u ;

[0029] 3) Initialize for t ∈ [1, T];

[0030] 4) Repeat;

[0031] 5)

[0032] 6) for t ∈ [1, T];

[0033] 7)

[0034] 8) Until convergence;

[0035] 9) Return θ f 、θ u 。

[0036] Furthermore, when the deep fuzzy cognitive map decouples the room temperature and humidity:

[0037] 1) Define the measured value and set value of the room humidity as the external factors of the deep fuzzy cognitive map, and the output value is the decoupling value of the room temperature;

[0038] 2) Define the measured value and the set value of the room temperature as the external factors of the deep fuzzy cognitive map, and the output value is the decoupled value of the room humidity.

[0039] Further, the RBF-PID algorithm in step four adjusts its own weight coefficients by the gradient descent method to provide the PID parameters: the proportional coefficient k p , the integral coefficient k i , the differential coefficient k d , and the learning rate η of the three coefficients pid .

[0040] Further, the control error in step four is e(k) = r(k) - y(k), where r(k) is the set value of temperature and humidity, and y(k) is the decoupled value of temperature and humidity.

[0041] Further, the output of the RBF-PID algorithm in step five is: where e(k - 1) is the control error at time k - 1.

[0042] Further, the transfer functions between the air volume controller and the air valve, and between the supply air temperature controller and the chilled water valve in step six are:

[0043]

[0044] where T is the sampling period, t is the delay time, and s is the variable of the transfer function obtained by the Laplace transform.

[0045] Compared with the prior art, the present invention has the following technical advantages:

[0046] (1) Based on the decoupling of the room temperature and humidity by the deep fuzzy cognitive map, the present invention realizes the real-time control of the room temperature and humidity in multiple regions by separately controlling the supply air volume and supply air temperature of the variable air volume air conditioning system. On the premise of ensuring the indoor thermal comfort environment, the RBF-PID algorithm calculates the supply air volume and supply air temperature at the end according to the decoupled value and the set value of the room temperature and humidity, and obtains the opening degrees of the end air valve and chilled water valve, so as to realize the real-time control of the temperature and humidity of the variable air volume air conditioning system in multiple regions and regulate the temperature and humidity of the rooms in multiple regions. This control method can improve the control accuracy of the end air valve and chilled water valve, and thus comprehensively improve the processing ability of the variable air volume air conditioning system in multiple regions in terms of dynamic interference. Compared with the prior art, the present invention solves the technical problems of poor control effect and weak anti-interference ability of the variable air volume air conditioning control system in multiple regions.

[0047] (2) The real-time control method of the present invention can solve the large lag problem existing in the room temperature and humidity regulation process of a multi-zone variable air volume air conditioning system. On the basis of ensuring the terminal control efficiency, it realizes the precise control of room temperature and humidity, which has important theoretical significance for the development of multi-zone variable air volume air conditioning systems and has practical application value and broad application prospects in engineering applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is the control block diagram of the real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system according to the present invention;

[0049] Figure 2 is the test platform diagram of a multi-zone variable air volume air conditioning system;

[0050] Figure 3 is the response curve diagram of the opening degree of the terminal air valve of a 3-room variable air volume air conditioning system to the room temperature;

[0051] Figure 4 is the response curve diagram of the opening degree of the chilled water valve of a 3-room variable air volume air conditioning system to the room humidity. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] In order to better understand the present invention, the following embodiments are used for illustration. The following embodiments belong to the protection scope of the present invention, but do not limit the protection scope of the present invention.

[0053] In the embodiment, as Figure 1 described, the real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system includes the following steps:

[0054] Step 1: Set the set values of room temperature and humidity;

[0055] Step 2: Obtain the temperature and humidity of the room through temperature and humidity sensors;

[0056] Step 3: Decouple the room temperature and humidity by using a deep fuzzy cognitive map;

[0057] In step 3, the deep fuzzy cognitive map can decouple the room temperature and humidity according to the potential relationship between room temperature and humidity:

[0058] 1) The state in the deep fuzzy cognitive map at time t is affected by the states of other concepts:

[0059]

[0060] Among them, the function f j (·) is used to simulate the relationship from a to a j and is called the f-function, Denote the system state of all concepts at time t with the function u j (·) to simulate the influence of external factors on a j , which is called the u-function;

[0061] 2) At the input layer of f j (a (t) ), assume At the output layer, an output can be obtained:

[0062]

[0063] Among them, is the output of the nth neuron in the Kth layer at time t, and υ(n1,K + 1) is the connection weight from the neurons in the Kth layer to the neurons in the K + 1th layer;

[0064] 3) The influence of external factors on a j can be indirectly measured by a u-function:

[0065]

[0066] 4) Assume Then the deep fuzzy cognitive map defined in can be rewritten as:

[0067]

[0068] Among them, is the decoupling value of , and θ f , θ u are the parameters of f j , u j respectively;

[0069] The said θ f , θ u are obtained by training the deep fuzzy cognitive map through the alternating function gradient descent algorithm:

[0070] 1) Input the training data set:

[0071] 2) Randomly initialize θ f , θ u ;

[0072] 3) Initialize For t ∈ [1, T];

[0073] 4) Repeat;

[0074] 5)

[0075] 6) For t ∈ [1, T];

[0076] 7)

[0077] 8) Until convergence;

[0078] 9) Return θ f and θ u ;

[0079] When the deep fuzzy cognitive map decouples the room temperature and humidity:

[0080] 1) Define the measured value and set value of the room humidity as the external factors of the deep fuzzy cognitive map, and the output value is the decoupled value of the room temperature;

[0081] 2) Define the measured value and set value of the room temperature as the external factors of the deep fuzzy cognitive map, and the output value is the decoupled value of the room humidity;

[0082] Step Four: Compare and subtract the decoupled values of the room temperature and humidity with the set values respectively. If the difference is not equal to 0, then use the difference as the control error e(k) and input it into the RBF-PID algorithm;

[0083] The RBF-PID algorithm adjusts its own weight coefficients using the gradient descent method to provide reasonable PID parameters: the proportional coefficient k p , the integral coefficient k i , the derivative coefficient k d , and the learning rate η of the three coefficients pid ;

[0084] The control error is e(k) = r(k) - y(k), where r(k) is the set value of the temperature and humidity, and y(k) is the decoupled value of the temperature and humidity;

[0085] Step Five: The RBF-PID algorithm calculates the supply air volume and supply air temperature at the next moment according to the difference, and obtains the opening degrees of the terminal air valve and the chilled water valve at the next moment;

[0086] The output of the RBF-PID algorithm is: where e(k - 1) is the control error at the (k - 1)th moment;

[0087] Step Six: Input the data of the opening degrees of the terminal air valve and the chilled water valve obtained by the RBF-PID algorithm into the air volume controller and the supply air temperature controller respectively, and the controllers control the terminal air valve and the chilled water valve to reach appropriate opening degrees;

[0088] The transfer functions between the air volume controller and the air valve, and between the supply air temperature controller and the chilled water valve are:

[0089]

[0090] Where T is the sampling period, t is the delay time, and s is the variable of the transfer function obtained by Laplace transform;

[0091] Step 7: Repeat Steps 2 to 6 in a loop to perform sampling at the sampling time interval.

[0092] To more clearly show the specific content and advantages of the present invention, a specific and detailed description of the actual specific implementation of the invention patent is provided.

[0093] The present invention proposes a real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system. First, a deep fuzzy cognitive map is used to decouple the room temperature and humidity. Then, the RBF-PID algorithm calculates the terminal air supply volume and supply air temperature based on the decoupled values and set values of the indoor temperature and humidity, and obtains the opening degrees of the terminal air valves and chilled water valves, thereby realizing the real-time control of the temperature and humidity of the variable air volume air conditioning system. This control method can improve the control accuracy of the terminal air valves and chilled water valves, and thus comprehensively improve the processing ability of the multi-zone variable air volume air conditioning system in terms of dynamic disturbances. To further illustrate the advantages of the present invention, an analysis is carried out in combination with a simulation example below. The specific implementation is as follows:

[0094] The RBF-PID algorithm adjusts its own weight coefficients using the gradient descent method, and provides reasonable PID parameters as follows: the proportional coefficient k p = 0.3, the integral coefficient k i = 0.35, the differential coefficient k d = 0.15, and the learning rate η pid = 0.15;

[0095] After the deep fuzzy cognitive map completes the decoupling of the room temperature and humidity, the RBF-PID algorithm can respond immediately. A random interference signal with an amplitude of 0.01 is added to the target output. The sampling period T = 60 seconds and the delay time t = 90 seconds. In the simulation, the transfer functions between the air volume controller and the air valve, and between the supply air temperature controller and the chilled water valve are:

[0096]

[0097] In this example, the variable air volume air conditioning system of the variable air volume air conditioning system experimental research platform is a single duct. The experimental platform is divided into 3 rooms and is located in a single-story building with a height of 4m, as Figure 2 shown. Table 1 shows the basic building information of each air-conditioned room, and Table 2 shows the main mechanical and electrical equipment of the VAV system.

[0098] Table 1 Basic building information of each air-conditioned room

[0099]

[0100] Table 2 Major mechanical and electrical equipment of the VAV system

[0101]

[0102]

[0103] To ensure the air quality in the room, this embodiment studies the response characteristics of the room temperature when the opening of the terminal air valve changes from 30% to 100%, and the response characteristics of the room humidity when the opening of the chilled water valve changes from 30% to 100%.

[0104] Figure 3 is the response curve of the opening of the terminal air valve of the variable air volume air conditioning system in 3 rooms to the room temperature, Figure 4 is the response curve of the opening of the chilled water valve of the variable air volume air conditioning system in 3 rooms to the room temperature.

[0105] From Figure 3 it can be seen that during 14:00 - 14:20, the opening of the terminal air valve remains at 30%, and the temperatures of the 3 rooms decrease slowly. At 14:20, the opening of the terminal air valve is manually adjusted from 30% to 100%. Around 14:24, the temperature curves of the 3 rooms show inflection points and the temperatures decrease significantly. At 14:40, the opening of the terminal air valve is manually adjusted from 100% to 30%. Around 14:44, the temperature curves of the 3 rooms show inflection points and the temperatures increase significantly. From Figure 4 it can be seen that during 14:00 - 14:20, the opening of the chilled water valve remains at 30%, and the humidities of the 3 rooms decrease slowly. At 14:20, the opening of the chilled water valve is manually adjusted from 30% to 100%. The humidity curves of the 3 rooms do not show obvious inflection points and the humidities continue to decrease. At 14:40, the opening of the chilled water valve is manually adjusted from 100% to 30%. Around 14:44, the humidity curves of the 3 rooms show inflection points and the humidities increase slowly. This indicates that the room temperature and humidity do not change immediately with the changes in the opening of the terminal air valve and the chilled water valve, and there will be a certain delay response time. After the opening of the terminal air valve and the chilled water valve is adjusted for 4 - 5 minutes, the indoor temperature and humidity will change significantly. Therefore, the delay response time of the indoor temperature and humidity to the opening of the terminal air valve and the chilled water valve is 4 - 5 minutes. That is, the control period can be selected as 5 minutes.

[0106] Based on the decoupling of room temperature and humidity by the deep fuzzy cognitive map, the RBF-PID algorithm calculates the terminal air supply volume and supply air temperature according to the decoupled values and set values of room temperature and humidity, obtains the opening degrees of the terminal air valve and the chilled water valve, and realizes the independent control of room temperature and humidity in multiple zones by separately controlling the air supply volume and supply air temperature of the variable air volume air conditioning system. It solves the large lag problem existing in the room temperature and humidity regulation process of the multi-zone variable air volume air conditioning system, realizes the precise control of room temperature and humidity on the basis of ensuring the terminal control efficiency, has important theoretical significance for the development of the multi-zone variable air volume air conditioning system, and has practical application value and broad application prospects in engineering applications.

[0107] Although the exemplary embodiments that are regarded as the present invention have been described and recounted, those skilled in the art will understand that various changes and substitutions can be made thereto without departing from the spirit of the present invention. Additionally, many modifications can be made to adapt a particular situation to the teachings of the present invention without departing from the central concept described herein. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but the present invention may also include all embodiments within the scope of the present invention and their equivalents.

Claims

1. A real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system, characterized in that, It includes the following steps: Step 1: Set the set values of the room temperature and humidity; Step 2: Obtain the room temperature and humidity through the temperature and humidity sensor; Step 3: Decouple the room temperature and humidity by using the deep fuzzy cognitive map; Step 4: Compare and subtract the decoupled values of the room temperature and humidity with the set values respectively. If the difference is not equal to 0, then use the difference as the control error e(k) and input it into the RBF-PID algorithm; Step 5: The RBF-PID algorithm calculates the terminal air supply volume and supply air temperature at the next moment according to the difference, and obtains the opening degrees of the terminal air valve and the chilled water valve at the next moment; Step 6: Input the opening degree data of the terminal air valve and the chilled water valve obtained by the RBF-PID algorithm into the air volume controller and the supply air temperature controller respectively, and the controller controls the terminal air valve and the chilled water valve to reach appropriate opening degrees; Step 7: Loop and execute Steps 2 to 6, and sample at the sampling time as the cycle; In Step 3, the deep fuzzy cognitive map can decouple the room temperature and humidity according to the potential relationship between the room temperature and humidity: 1) States in the Deep Fuzzy Cognitive Map Affected by the states of other concepts at time t: Among them, the function f j (·) is used to simulate the relationship from a to a j , which is called the f-function, represents the system state of all concepts at time t, and the function u j (·) is used to simulate the influence of external factors on a j , which is called the u-function; 2) At f j (a (t) ) of the input layer, set At the output layer, an output can be obtained: Among them, is the output of the n-th neuron in the K-th layer at time t, and v(n1, K+1) is the connection weight from the neuron in the K-th layer to the neuron in the K+1-th layer; 3) The influence of external factors on a j can be indirectly measured by a u-function: 4) Let Then The deep fuzzy cognitive map defined in Among them, is the decoupling value, θ f , θ u are respectively the parameters of f j , u j ; The said θ f and θ u are obtained by training a deep fuzzy cognitive map through an alternating function gradient descent algorithm: 1) Input the training data set: 2) Randomly initialize θ f , θ u ; 3) Initialization For t ∈ [1, T]; 4) Repeat; 5) 6) For t ∈ [1, T]; 7) 8) Until convergence; 9) Return θ f , θ u ; When the deep fuzzy cognitive map decouples the room temperature and humidity: 1) Define the measured value and set value of the room humidity as the external factors of the deep fuzzy cognitive map, and the output value is the decoupled value of the room temperature; 2) Define the measured value and set value of the room temperature as the external factors of the deep fuzzy cognitive map, and the output value is the decoupled value of the room humidity.

2. The real-time control method for temperature and humidity of a multi-zone variable air volume air conditioning system according to claim 1, characterized in that, In step 4, the RBF-PID algorithm adjusts its own weight coefficients using the gradient descent method to provide PID parameters: the proportional coefficient k p , the integral coefficient k i , the differential coefficient k d , and the learning rate η of the three coefficients pid .

3. The real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system according to claim 1, characterized in that, The control error in Step 4 is e(k) = r(k) - y(k), where r(k) is the set value of the temperature and humidity, and y(k) is the decoupled value of the temperature and humidity.

4. The real-time control method for the temperature and humidity of a multi-zone variable air volume air conditioning system according to claim 1, wherein, The output of the RBF-PID algorithm described in step five is: where e(k - 1) is the control error at time k - 1.

5. The control method for temperature and humidity of a multi-zone variable air volume air conditioning system based on a deep fuzzy cognitive map according to claim 1, characterized in that, The transfer functions between the air volume controller and the air valve, and between the supply air temperature controller and the chilled water valve in Step 6 are: Where T is the sampling period, t is the delay time, and S is the variable of the transfer function obtained by the Laplace transform.

Images